Small Modular Reactors

Every operating commercial reactor in the US produces 1,000+ MW. SMRs propose reactors at 50-300 MW, factory-built in modules and shipped to site. The theory: smaller units reduce upfront capital risk, enable factory learning curves, and fit where large reactors cannot. The practice: no commercial SMR operates in the US, and the first attempt (NuScale UAMPS) was cancelled.

The economic case for SMRs rests on manufacturing, not physics. Large reactors are custom-built on-site, each effectively a prototype. Construction takes 7-12 years, costs $10-15 billion, and routinely exceeds budgets. SMRs aim to shift construction from the field to the factory, where controlled conditions enable standardization, quality control, and learning-curve cost reductions. Instead of one $12 billion project, a utility deploys four $800 million modules over time, reducing financial risk.

The unproven assumption. SMR proponents cite the learning curve that made solar cheap. But solar modules are identical units produced in millions. SMR factories might produce dozens per year, possibly too few to achieve meaningful learning rates. The per-MW capital cost of first-of-a-kind SMRs is estimated at $8,000-15,000/kW, roughly 2-3x a large reactor's per-MW cost. Cost competitiveness requires production volumes that no market has yet demanded.

Can SMRs compete with solar and batteries on cost?

Not on LCOE today. At $8,000-15,000/kW, SMR electricity would cost $90-180/MWh versus $30-50/MWh for solar-plus-storage. SMRs compete on different attributes: firm capacity, small footprint, and load-following capability. Whether these attributes command a sufficient premium is the commercial question that NuScale's cancellation left unanswered.

---